Fig 1.
Body-connected routing scenario.
Fig 2.
Principle of transfer learning.
Fig 3.
VGR structure.
Fig 4.
Comparison of normalized throughput and load.
Fig 5.
Average bandwidth utilization and load comparison chart.
Fig 6.
Transmission time under different transmission numbers.
Fig 7.
Normalized transmission delay with time.
Q-learning is similar to RL, but its disadvantage is that it is not good at handling complex data. As shown in the figure, Q-learning has a higher transmission delay than RL.
Fig 8.
Normalized convergence time.